sharp s (Eszett)

[Martin Duerst]

At 08:01 08/03/08, Mark Davis wrote:
>The main reason for mapping <sharp s> to "ss" in IDNA2003 is for case insensitivity.

Of course. But that does not fully explain IDNA2003. There are many
other cases where case insensitivity would have lead to different
design decisions. As an example, take French. Applying the same
case folding principles would have removed all accents.

The main reason for IDNA2003 being the way it is is that the IETF
needed a case-folding table, and the Unicode Consortium had a
(overall very good) one available. The problem was that said
case-folding table was designed for a slightly different job
than IDNA, and that the IETF was affraid to change it because
they didn't want to open the door for dozens of other (mostly
case-unrelated) change requests.


>That is, so that "<http://Ru$B%F](Bcom>Ru<sharp s>.com" matches "<http://RUSS.COM>RUSS.COM", which in turn matches "<http://russ.com>russ.com". A small side benefit is that some words that are spelled differently by Swiss, or differently pre/post Spelling Reform also have the same internal form, but that side benefit was really given no weight in the original decision. Remember, also, that a great many words of every language cannot be simply used as is with IDNA. One can't use many extremely common English words; for example, "can'<http://t.com>t.com" won't work.
>
>At this point, I really doubt that that the advantage of having <sharp s> outweighs the cost of either incompatibility with IDNA2003, or the morass that would be caused by a prefix change.

In my opinion, definitely not the later, but I don't think that's necessary.
The cost of the incompatibility with IDNA2003 may well be worth it. It would
mainly be paid by registrants (additional cost for registring both variants)
and registrars/registries (costs for bundling,...).

Regards,   Martin.



#-#-#  Martin J. Du"rst, Assoc. Professor, Aoyama Gakuin University
#-#-#  http://www.sw.it.aoyama.ac.jp       mailto:duerst at it.aoyama.ac.jp